GPU Frequency Control via Command Parsing for Power Management

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing power management techniques for graphics processing units (GPUs) lead to increased power consumption due to performance improvements, and sampling-based methods result in delayed responses to fluctuating workloads.

Innovation Solution

A system and method for predicting future GPU operation times based on command information and historical data to adjust GPU frequency proactively, ensuring efficient power management by parsing commands, determining operation times, and adjusting frequency accordingly.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Use of energy by stationary object

If sampling-based power management is used to monitor GPU workload, then power consumption can be managed, but the response to fluctuating workloads is delayed

Engineering Contradiction:

Improvepower consumptionVSAvoidresponse delay

Core Design Contradiction:

Use of energy by stationary objectVSLoss of time

Solution Approach 1:

The patent parses command buffers to predict future GPU workload before it executes, allowing frequency adjustment to be made in advance based on anticipated demand rather than reacting to past workload samples. This predictive approach eliminates the inherent delay in sampling-based methods by acting beforehand on forecasted conditions.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements a feedback loop where command buffer parsing continuously monitors upcoming workload characteristics, predicts operation times, and adjusts frequency accordingly. This closed-loop control ensures power management responds dynamically to actual workload fluctuations without delay.

Inventive Principle:

Principle #23Feedback

2Productivity

If GPU frequency is increased to improve performance, then processing speed increases, but power consumption increases

Engineering Contradiction:

Improveprocessing speedVSAvoidpower consumption

Core Design Contradiction:

ProductivityVSUse of energy by stationary object

Solution Approach 1:

The patent dynamically adjusts GPU frequency based on predicted workload characteristics by parsing command buffers. The frequency control component modifies operating frequency in real-time according to anticipated operation times and workload intensity, optimizing the balance between performance and power consumption rather than using static or oversampled adjustments.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes the operating frequency parameter of the GPU based on parsed command information and predicted operation times. By adjusting this key parameter dynamically according to actual workload demands rather than fixed sampling intervals, the system achieves efficient power-performance trade-offs.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS9875516B2Systems and methods for graphics process units power management

Publication Date: 2018.01.23 MARVELL ASIA PTE LTD

AI summary

Systems and methods are provided for frequency adjustment of graphics process units (GPUs). A system includes: a command parser configured to parse one or more first commands associated with one or more future GPU operations to obtain command information, a processing component configured to determine an operation time for the future GPU operations based at least in part on the command information, and a frequency control component configured to adjust a GPU frequency based at least in part on the operation time for the future GPU operations.